CN112350441A

Movatterモバイル変換

Info

Publication number: CN112350441A
Application number: CN202011212251.8A
Authority: CN
Inventors: 杨森; 孙志周; 田克超; 裴淼; 贾同辉; 刘强; 赵学强; 李北斗; 孙凯; 魏德凯; 刘加科; 杨国庆; 陈姣; 袁立国
Original assignee: State Grid Intelligent Technology Co Ltd
Current assignee: State Grid Intelligent Technology Co Ltd
Priority date: 2020-11-03
Filing date: 2020-11-03
Publication date: 2021-02-09
Anticipated expiration: 2040-11-03
Also published as: CN112350441B; WO2022095616A1

Abstract

Translated fromChinese

本发明提供了一种变电站在线智能巡视系统及方法，获取主辅设备监控系统的联动信号，能够做到融合多维信息；同时，利用不同执行系统并发执行相应子任务，保证了巡检效率；基于各个执行系统所采集数据，融合了多维度检测数据，并进行全自动智能识别、分析，形成联动复核；本发明能够实现变电站的自动化全覆盖在线巡视，异常故障的自动复核处置，运维人员远程运维管理，解决了机器人巡检存在巡视盲区、单一视频覆盖范围小、故障应急处置不及时等问题，提高了变电站设备巡检的全面性、准确性、智能性，提升了变电站内系统联动能力。

The invention provides an online intelligent inspection system and method for a substation, which can obtain linkage signals of a monitoring system of main and auxiliary equipment, and can integrate multi-dimensional information; at the same time, different execution systems are used to execute corresponding subtasks concurrently, thereby ensuring inspection efficiency; The data collected by each execution system integrates multi-dimensional detection data, and performs automatic intelligent identification and analysis to form a linkage review; the invention can realize automatic full coverage online inspection of substations, automatic review and disposal of abnormal faults, and remote operation and maintenance personnel. Operation and maintenance management solves the problems of robot inspection blind spots, small coverage of a single video, and untimely emergency response to faults, improves the comprehensiveness, accuracy, and intelligence of substation equipment inspections, and improves system linkage capabilities in substations .

Description

Online intelligent inspection system and method for transformer substation

Technical Field

The invention belongs to the technical field of intelligent inspection of power systems, and particularly relates to an online intelligent inspection system and method for a transformer substation.

Background

The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.

At present, a plurality of transformer substations gradually replace manual inspection work with intelligent robots and video acquisition equipment, great convenience is provided for intelligent inspection of the transformer substations, and inspection effect and quality are also improved.

However, according to the inventor's knowledge, the inspection method has the following technical problems:

(1) the inspection means of the inspection robot is only to detect the equipment parameters through a sensor, and does not analyze other dimension detection references;

(2) only the robot and the video information are acquired, the acquired information source is single, the association with other important information (such as linkage signals of a main and auxiliary equipment monitoring system) cannot be realized, the defect of data cooperative processing is great, meanwhile, the alarm rechecking of the acquired information is lacked, and the accuracy is poor;

(3) the inspection robot is only depended on to universally have inspection blind areas when the inspection is carried out, and the full coverage of the inspection cannot be realized.

Disclosure of Invention

The invention provides a transformer substation online intelligent inspection system and a transformer substation online intelligent inspection method, which can realize automatic full-coverage online inspection of a transformer substation, automatic rechecking and disposal of abnormal faults and remote operation and maintenance management of operation and maintenance personnel, solve the problems of inspection blind areas, small single video coverage range, untimely fault emergency disposal and the like in robot inspection, improve the comprehensiveness, accuracy and intelligence of transformer substation equipment inspection, and improve the system linkage capacity in the transformer substation.

According to some embodiments, the invention adopts the following technical scheme:

a transformer substation online intelligent inspection method comprises the following steps:

acquiring a linkage signal of main and auxiliary equipment, acquiring a related equipment point location list according to the linkage signal, and constructing an inspection task or receiving the inspection task according to the equipment point location and a pre-configured linkage strategy;

dividing the obtained inspection task into a plurality of subtasks, and respectively executing the corresponding subtasks by different inspection execution systems in parallel;

concurrently executing the data acquired by the corresponding subtasks by each inspection execution system to form multi-dimensional inspection data;

and determining an intelligent analysis algorithm adopted by the acquired data by using a cloud edge cooperation method, and identifying the equipment state by using the intelligent analysis algorithm to obtain an identification result.

In the invention, the inspection task receiving means that an inspection task scheme formed according to user requirements and customization can be received.

According to the scheme, the method and the device can acquire the linkage signals of the main and auxiliary equipment monitoring systems, and can fuse multidimensional information; meanwhile, different execution systems are utilized to execute corresponding subtasks concurrently, so that the inspection efficiency is ensured; based on the data collected by each execution system, multi-dimensional detection data are fused, full-automatic intelligent recognition and analysis are carried out, and linkage recheck is formed.

As an alternative implementation, the specific process of acquiring the linkage signal of the main and auxiliary devices and acquiring the associated device point location list according to the linkage signal includes:

finding out a related equipment point location list according to a main equipment code of a linkage signal of the main and auxiliary equipment monitoring systems;

acquiring a related linkage strategy list from linkage strategy configuration according to the equipment point location list;

judging whether the linkage strategy list is empty, and if not, constructing a substation equipment joint inspection task according to the equipment point location list and the linkage strategy list; and if the linkage strategy list is empty, the linkage is finished.

In alternative embodiments, the inspection execution system includes, but is not limited to, an optical inspection system, a robot inspection system, an environment monitoring system, and a sensing parameter acquisition system.

As an alternative embodiment, the specific method for the optical patrol system to perform the corresponding subtasks includes: analyzing the content of the corresponding subtask, extracting target equipment to be detected, and calling optical equipment within a certain range from the target equipment to be detected to acquire videos, images or/and spectrums of the target equipment to be detected.

As an alternative embodiment, the specific method for the environment monitoring system to execute the corresponding subtask includes: and the environment monitoring system analyzes the corresponding subtasks, and controls and adjusts the illumination and/or temperature in the transformer substation according to the subtask content.

As an alternative embodiment, the specific method for the robot inspection system to execute the corresponding subtasks includes: analyzing the content of the corresponding subtask, extracting target equipment to be detected and the inspection point, controlling the robot within a certain range from the target equipment to move to the corresponding inspection point to execute linkage action according to a linkage strategy associated with the equipment point position, and associating the target equipment with inspection data and an inspection result.

As alternative embodiments, the sensing parameter collecting system includes, but is not limited to, several of a voiceprint detecting device, a visible light collecting device, an infrared collecting device, an ultraviolet collecting device, and a sound collecting device.

By way of further limitation, the robot performs a process comprising: judging whether the robot switches indoor and outdoor environments, if so, opening an automatic door of the protection room, and closing the automatic door of the protection room after the robot enters or leaves the protection room;

if the current equipment point location is outdoor equipment, judging whether rain and snow exist at present, and if so, judging whether the current rainfall and snowfall exceed the threshold borne by the robot according to the rainfall and snowfall;

if the rainfall and the snowfall exceed the threshold born by the robot, routing inspection is carried out by using the optical equipment associated with the current equipment point location, the optical equipment is controlled to acquire routing inspection data of the equipment point location, data identification is carried out, a routing inspection result is obtained, and analysis is carried out according to the routing inspection data and the routing inspection result;

and calculating a route for the robot to return to the charging room according to the current position of the robot and the position of the charging point, and controlling the robot to move according to the route.

As a further limitation, if the rainfall and snowfall do not exceed the threshold values borne by the robot, controlling the robot to perform a wiper operation;

and controlling the robot to acquire the inspection data of the equipment point location, performing data identification, acquiring an inspection result, and completing early warning analysis according to the inspection data and the inspection result.

By way of further limitation, the patrol point is obtained by pre-configuration, and the configuration method comprises the following steps: acquiring a three-dimensional model of a polling place, and preprocessing the three-dimensional model;

configuring a reference coordinate system for the preprocessed model, identifying a target object in the inspection site, acquiring pose and size information of the target object, and extracting a passable path in the inspection site;

configuring the constraint conditions of the distance range between the inspection point and the target object and the plane normal deviation angle range of the target object;

traversing the target object, and calculating a legal configuration area according to the pose and the size of the single target object and the passable path information;

carrying out global optimization on a legal configuration area of a single target object, and solving the coordinates of an optimal inspection point under a constraint condition aiming at a public crossing area and a non-crossing area to obtain inspection point information;

and calculating the corresponding three-dimensional posture of the robot according to each target object and the corresponding optimal three-dimensional coordinates of the inspection points to obtain an inspection point information list of the robot.

As an alternative embodiment, the specific process of determining the intelligent analysis algorithm by using cloud edge cooperation includes:

loading an intelligent analysis algorithm to be updated, constructing and authenticating identity;

if the authentication information is legal, the algorithm data is divided into packets and the data is encrypted;

and decrypting the algorithm data packets, verifying the usability of the algorithm after all the algorithm data packets are received, taking out the old algorithm in the intelligent analysis algorithm bin according to the type of the algorithm application object, performing compression backup, and adding the algorithm to be updated into the intelligent analysis algorithm bin.

As an alternative embodiment, the specific process of performing the equipment fault diagnosis and analysis by using the intelligent analysis algorithm is as follows: and comprehensively identifying the running state of the equipment based on the multi-dimensional routing inspection data, and determining that the equipment is in fault when at least one information analysis result is in fault.

In the present invention, the intelligent analysis algorithm may be a deep learning model or an artificial intelligence algorithm.

As an alternative embodiment, the multi-dimensional inspection data includes, but is not limited to, video information, spectral information, target device appearance data, and target device sound information.

As a further limitation, the specific process of performing state identification according to the sound information of the target device includes: acquiring sound information of target equipment of a transformer substation, and preprocessing the sound information;

extracting voiceprint characteristics of the preprocessed voice information; wherein the voiceprint features at least comprise FBank features, decibels, fundamental frequencies, short-time energy, short-time zero crossing rates and correlation coefficients;

inputting the extracted voiceprint characteristics into a trained voiceprint recognition model, and outputting a recognition result;

and judging the operation state of the target equipment based on the identification result.

An online intelligent patrol system of a transformer substation comprises a patrol processing system, and an optical patrol system, a robot patrol system, an environment monitoring system and a sensing parameter acquisition system which are connected with the patrol processing system;

the inspection processing system is in linkage interaction with the main and auxiliary monitoring systems and is configured to acquire main and auxiliary equipment linkage signals, acquire a related equipment point location list according to the linkage signals, and establish an inspection task according to the equipment point location and a pre-configured linkage strategy or establish the inspection task according to requirements;

the inspection processing system divides the inspection task into a plurality of subtasks according to the inspection task, and the optical inspection system, the robot inspection system, the environment monitoring system and the sensing parameter acquisition system respectively execute the corresponding subtasks in parallel;

the patrol processing system is configured to determine an adopted intelligent analysis algorithm by utilizing a cloud-edge cooperation method according to video data, equipment patrol data, environment parameter data and sensing parameter data which are acquired by concurrently executing corresponding subtasks by each system, analyze the data by utilizing the intelligent analysis algorithm, identify the state of equipment, and give an alarm or/and operate and maintain the equipment according to an identification result;

the tour processing system is further configured to provide a video analysis service, a spectral analysis service, a voiceprint analysis service, and a pattern recognition service.

As an alternative embodiment, the patrol processing system can be extended to cloud deployment.

As an alternative embodiment, the optical inspection system includes a plurality of optical devices disposed in the station end, the optical devices include several kinds of fixedly mounted visible light guns, visible light ball machines, thermal infrared imagers, dual spectrum ball machines and dual spectrum cloud platforms, and the optical devices are all connected to the data collection and exchange device.

As an alternative embodiment, the robot inspection system includes a plurality of indoor and outdoor robots, and the indoor and outdoor robots are connected to the data collection and exchange device by wireless or wired means.

Of course, the robot can select a wheeled robot, a track robot, a tracked robot and the like according to a specific use scene.

As an alternative embodiment, the environment monitoring system comprises an illumination controller, a curtain controller, a temperature controller and a plurality of temperature sensors and humidity sensors, wherein the illumination controller is used for turning on or off the lighting device, the curtain controller is used for turning on or off the curtain, the temperature sensors and the humidity sensors are used for acquiring the temperature and the humidity in the station, and the temperature controller is used for controlling the temperature adjusting mechanism.

The temperature adjusting mechanism includes but is not limited to an air conditioner and a heater.

As an alternative embodiment, a plurality of voiceprint acquisition devices are further arranged in the station end, and the voiceprint acquisition devices are arranged on the robot or beside the power equipment.

A computer readable storage medium, having stored therein a plurality of instructions adapted to be loaded by a processor of a terminal device and to perform the steps of a substation online intelligent patrol method as described above.

A terminal device comprising a processor and a computer readable storage medium, the processor being configured to implement instructions; the computer readable storage medium is used for storing a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the steps of the transformer substation online intelligent patrol method.

Compared with the prior art, the invention has the beneficial effects that:

the invention innovatively provides an online intelligent inspection method for a transformer substation, which is characterized in that inspection tasks are analyzed, and each inspection execution system concurrently executes subtasks, so that automatic inspection, real-time monitoring, intelligent linkage and remote control are realized, the application range of online intelligent inspection of the transformer substation is expanded, the informatization and automation levels of the inspection process and the inspection result processing of the transformer substation are improved, the inspection capability and the equipment defect remote processing capability of the transformer substation are enhanced, the labor cost is reduced, and the unattended process of the transformer substation is promoted.

The invention innovatively provides a multi-dimensional linkage rechecking method for inspection data, which utilizes inspection execution systems such as robots, optical equipment, voiceprint devices and the like to carry out combined inspection, fully-automatic intelligent acquisition, identification and analysis, responds to multi-dimensional linkage rechecking of substation equipment alarm information, improves the comprehensiveness and intelligence of inspection, and ensures the alarm accuracy.

The invention innovatively provides a cloud-side cooperation method, and by means of standardized access, automatic download updating and scene iteration of intelligent analysis algorithms such as a deep learning model and an artificial intelligence algorithm, the hardware cost of an edge side is effectively saved, the delay of edge side data is reduced, the application effect of the edge side is automatically improved, and the effectiveness of cloud side data and the resource utilization rate are improved.

The invention innovatively provides a robot system, an optical monitoring system and an environment monitoring system linkage control method, external environment parameters such as illumination and airflow are effectively controlled, the interference of environment on robot inspection is reduced, the robot inspection adaptability of different inspection scenes is improved, the robot inspection range is expanded, and the reliability and the inspection data accuracy of robot inspection are improved.

The invention innovatively provides a three-dimensional inspection tour method for fusing a three-dimensional model of a transformer substation with videos, which adopts a three-dimensional panoramic preview and immersive roaming mode of the transformer substation to realize scene positioning, information viewing inspection resource coverage analysis and uncovered inspection point positions to provide a newly-added video point arrangement strategy; the problem that the traditional inspection visual angle is single is changed, the equipment is subjected to omnibearing three-dimensional inspection, the health condition of the equipment is better mastered, the fault position and the relevant condition are more accurately determined, and a positive effect is achieved on reasonably making an inspection plan.

The invention innovatively provides a transformer substation equipment voiceprint monitoring method, a sound collection system is adopted to analyze equipment sound by utilizing a voiceprint recognition technology of voiceprint feature extraction, voiceprint recognition and model classification, so that equipment sound abnormity recognition is realized, sound collection precision and recognition accuracy are improved, and technical support is provided for intelligent audio recognition of the running state of power equipment. The problem of transformer substation's sound collection positioning accuracy is poor and the frequency domain span is big is solved, the collection of many first directional sound signals has been realized, the interference of outdoor environmental noise and artificial noise has been avoided, the sound signal that single microphone was gathered is the stack of diversified sound signal, can realize noise suppression and sound source location, has improved the rate of accuracy of sound collection and discernment.

In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the invention and together with the description serve to explain the invention and not to limit the invention.

FIG. 1 is a schematic structural diagram of an online intelligent inspection system of a transformer substation;

FIG. 2 is a schematic view of a tour system;

FIG. 3 is a schematic diagram of a robotic system;

FIG. 4 is a schematic diagram of an environment monitoring system;

FIG. 5 is a schematic flow chart of an online intelligent inspection method for a transformer substation;

FIG. 6 is a schematic diagram of a task flow of online intelligent inspection of a transformer substation;

FIG. 7 is a schematic flow chart of a cloud-side coordination method for online intelligent patrol of a transformer substation;

FIG. 8 is a schematic diagram of the primary and secondary linkages;

FIG. 9 is a schematic view of an environmental system linkage;

FIG. 10 is a schematic diagram of an automatic configuration process of inspection points in an inspection task;

FIG. 11 is a schematic diagram of the operation and maintenance of the apparatus;

fig. 12 is a schematic diagram of a device status identification process based on voiceprint detection.

The specific implementation mode is as follows:

the invention is further described with reference to the following figures and examples.

It is to be understood that the following detailed description is exemplary and is intended to provide further explanation of the invention as claimed. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.

The first embodiment is as follows: as shown in fig. 1, the present embodiment provides an online intelligent substation inspection system, which includes an inspection processing system, an optical inspection system, a robot inspection system, an environment monitoring system, and a sensing parameter acquisition system, where the inspection processing system is connected to the optical inspection system, the robot inspection system, the environment monitoring system, and the sensing parameter acquisition system, and is connected to a main and auxiliary equipment monitoring system through a data collection and exchange device;

the inspection processing system is in linkage interaction with the main and auxiliary monitoring systems and is configured to acquire main and auxiliary equipment linkage signals, acquire a related equipment point location list according to the linkage signals, construct an inspection task according to the equipment point location and a pre-configured linkage strategy or receive an input inspection task customized according to user requirements;

the patrol processing system is configured to determine an adopted intelligent analysis algorithm by using a cloud-edge cooperation method according to video data, equipment patrol data, environment parameter data and voiceprint data which are acquired by concurrently executing corresponding subtasks by each system, analyze the data by using the intelligent analysis algorithm, identify the state of the equipment, and give an alarm or/and operate and maintain the equipment according to an identification result.

The system also comprises an image storage device, a network device, a plurality of robots, a plurality of optical devices, a plurality of voiceprint devices and the like. The devices may all be accessible through the access stratum. The access layer provides two modes of wireless security and wired security access. The access stratum may include secure access devices. Of course, the access stratum may be connected to other devices, such as an acquisition module for other parameters, which is not exhaustive herein.

The following are detailed separately:

the patrol processing system is connected with the data collection and exchange equipment. The data collection and exchange device is used as a hub of network connection and is connected with network devices among systems. The main and auxiliary equipment linkage signals are sent to the data collection exchange equipment through the network safety protection equipment, and linkage results and reverse linkage signals are sent back to the main and auxiliary equipment monitoring system through the network safety protection equipment.

In this embodiment, the network security protection device may select a forward isolation device, a reverse isolation device, or other devices according to the situation.

The image storage device can be a network hard disk video recorder.

The environment monitoring system is connected with the data collection and exchange equipment. The environment monitoring system includes an environment parameter detection module, such as a temperature detection module, a humidity detection module, etc., which are not described herein again. And the system also comprises or is connected with an environmental parameter control execution mechanism, and as shown in fig. 4, the system can comprise an environmental monitoring management machine, a protocol conversion device, a temperature and humidity sensor, an air speed sensor, a rain and snow sensor, a gas sensor, an automatic door controller, an illumination controller, a curtain controller and an air conditioner controller. (details will be described in the context of the specific environment linkage control section).

The optical devices comprise a visible light gun (gun type network camera is abbreviated as gun), a visible light ball machine (ball type network camera is abbreviated as ball machine), an infrared thermal imager, a double-spectrum (simultaneously having visible light and infrared thermal imaging and being called double-spectrum) ball machine and a double-spectrum holder which are fixedly installed. The optical devices are converged by the video converging and storing and forwarding device and then communicated with the image storage device, and the image storage device is also connected with the data converging and switching device.

In this embodiment, the video aggregation, storage and forwarding device may be a switch, and the data aggregation and switching device may be a router, which is responsible for protocol conversion and data exchange. The protocol conversion device may employ a serial server.

The robot system comprises a plurality of outdoor robots and a plurality of indoor robots. The outdoor robot is connected to the data collection exchange equipment through a wireless network, in order to guarantee the safety of wireless network access, the outdoor robot is firstly connected to a safety access module, the safety access module is connected to a wireless network bridge Station, the wireless network bridge Station is connected to a wireless network bridge AP through the wireless network, the wireless network bridge AP is connected to a safety access platform, and the safety access platform is connected to the data collection exchange equipment. The security access platform is used in a matching way with the security access module, and the security of the wireless network is ensured through physical isolation, data encryption and identity authentication functions. The indoor robot may also be connected to the data collection and exchange device via a wired network.

Of course, the robot can select a wheeled robot, a track robot, a tracked robot and the like according to a specific use scene. And will not be described in detail herein.

As shown in fig. 2, the patrol processing system is provided including: the system comprises an online intelligent patrol service, a user interface, a configuration tool, a video execution unit, a pattern recognition service, a deep learning service, a video analysis service and a voiceprint analysis service; respectively comprises the functional modules.

The patrol processing system can also be connected with or communicated with an upper cloud end server or a centralized control server. Or deployed in the cloud. In this embodiment, the patrol processing system may include a patrol host and software configured thereon.

The online intelligent patrol service is a core business service of a patrol processing system, and comprises 11 functional modules of message distribution, model management, task management, intelligent linkage, a communication interface, real-time monitoring, data analysis, three-dimensional application, cloud-edge cooperation, alarm analysis and configuration management.

The message distribution module is responsible for message circulation among the functional modules of the online intelligent patrol service, and other functional modules can realize the transmission of communication messages among the functional modules through the subscription and the release of the messages.

The model management module is responsible for managing a transformer substation equipment model, a transformer substation equipment point location model, a robot model, an optical equipment model and an online monitoring model, synchronizing transformer substation equipment and equipment point locations online and binding with the transformer substation equipment point location model configured by the robot.

The task management is responsible for task model management, task customization process, task timing process, task control, task state monitoring, routing inspection result access and storage, and task report generation and pushing.

The intelligent linkage module is responsible for linkage strategy configuration, linkage signal monitoring, linkage judgment, linkage action execution and linkage execution information pushing.

The communication interface module is responsible for communicating with the user interface, the configuration tool and the video execution unit.

The real-time monitoring module is responsible for real-time data access of the robot, real-time data access of optical equipment, on-line monitoring real-time data access, peripheral data access of the robot, historical data cleaning and real-time data pushing.

The data analysis module is responsible for communicating with the pattern recognition service, the deep learning service, the video analysis service and the voiceprint analysis service, and completes analysis of the inspection data to form an inspection result.

The three-dimensional application module is responsible for providing functions of three-dimensional panoramic browsing and immersive roaming display, alarm positioning, video pushing, space ranging, resource coverage analysis and the like.

The cloud edge coordination module is responsible for synchronous updating of the analysis algorithm and the application of the advanced analysis service by the intelligent analysis cloud platform.

The alarm analysis module is responsible for alarm threshold setting, alarm analysis, alarm rechecking, alarm information pushing, alarm auditing and defect management.

The configuration management module is responsible for providing functions of alarm threshold setting, alarm message subscription setting, authority management, standard point location library maintenance and the like for a user.

The user interface is realized by a Web system and is used for providing real-time video and state display of the robot and the optical equipment, robot and equipment control, routing inspection task customization and control, routing inspection task progress and state, routing inspection task report forms, routing inspection result real-time display, alarm information pushing, routing inspection result inquiry and verification, alarm inquiry and verification, map and three-dimensional information display for a user.

The configuration tool is responsible for completing system configuration, model configuration, database configuration, file service configuration and loading module configuration of the online intelligent patrol service.

The video task execution unit is responsible for task execution flow, polling data acquisition and video monitoring of the optical equipment, and completes data identification and analysis of the visible light picture and the infrared heat map through communication with the mode identification service.

The mode identification service module is responsible for identifying the disconnecting link on-off state, the meter reading, the infrared temperature, the light, the pressing plate, the two-dimensional code and the foreign matter picture.

The deep learning service module is responsible for identifying appearance defects of the substation equipment, identifying state defects of the substation equipment and judging abnormal changes of the substation equipment.

The video analysis service module is responsible for identifying the safety risk defects in real time and comprises the following steps: crossing/intruding, not wearing safety helmet, not wearing frock, smoking, firework identification, toy identification, ponding monitoring.

The voiceprint analysis service module is responsible for extracting characteristic quantities from audio signals generated by vibration of the substation equipment and identifying the running state of the equipment.

As shown in fig. 3, the robot system includes a robot execution unit, a motion control unit, a positioning navigation service unit, a pattern recognition service unit, a path planning service unit, and a multi-robot communication unit. The robot execution unit is responsible for task execution flow of the robot, polling data acquisition, video monitoring, robot body alarming, autonomous navigation and external communication.

The motion control unit is responsible for pan-tilt control, power control, drive control, automatic door control, elevator control and control state uploading of the robot. The positioning navigation service is responsible for providing a laser positioning navigation function, an indoor track positioning navigation function and a GPS positioning navigation function.

The pattern recognition service unit has the same function as the pattern recognition service in the patrol processing system.

And the path planning service unit is responsible for loading the map model data of the transformer substation, loading a path planning algorithm and providing a path planning service.

The multi-robot communication unit is responsible for completing the communication function among the multiple robots, and through publishing and subscribing of messages, sharing of position information, environment information, electric quantity information, task state information and motion state information among the multiple robots is completed.

Example two: the method for the online intelligent inspection of the transformer substation comprises the following steps:

acquiring a linkage signal of main and auxiliary equipment, acquiring a related equipment point location list according to the linkage signal, and constructing an inspection task or receiving the inspection task according to the equipment point location and a pre-configured linkage strategy (of course, the inspection task can be customized according to the requirement);

acquiring a polling task, splitting the polling task into a plurality of subtasks, and respectively executing the corresponding subtasks by an optical polling system, a robot polling system and an environment monitoring system in parallel;

concurrently executing video data, equipment inspection data, environmental parameter data and voiceprint data acquired by corresponding subtasks by each system;

and determining an intelligent analysis algorithm adopted by the acquired data by using a cloud edge cooperation method, and identifying the equipment state by using the intelligent analysis algorithm based on the acquired data to obtain an identification result.

Specifically, as shown in fig. 5, the method includes the following steps:

and a polling task generation step: acquiring a linkage signal of main and auxiliary equipment, acquiring a related equipment point location list according to the linkage signal, and constructing an inspection task according to the equipment point location and a pre-configured linkage strategy;

of course, in this embodiment, the polling task is not triggered only by the linkage signal, but also can be customized to create a polling scheme.

Before the routing inspection task is constructed, the use authority of the system is managed, whether the system is an operator with authority is determined, and if the authority management is passed, routing inspection scheduling is performed.

And the polling scheduling subsystem is linked with the main and auxiliary equipment monitoring systems according to the authorization and the task sequence of the user.

A data acquisition step: according to the inspection task, a task control system of the online intelligent inspection system of the transformer substation divides the inspection task into a plurality of subtasks, and the optical inspection system, the robot inspection system, the environment monitoring system and the sensing parameter acquisition system (including various sensors or other acquisition modules arranged in a station, such as visible light, infrared light, ultraviolet light, sound, voiceprint and other parameter acquisition devices) concurrently execute the corresponding subtasks respectively;

the method comprises the steps that a station end transfers video data, equipment inspection data (including but not limited to appearance images and equipment temperature data), environment parameter data (including but not limited to temperature, humidity and illumination) and sound pattern data collected by all systems for concurrently executing corresponding subtasks to a real-time database for storage, determines an intelligent analysis algorithm of the station end by using a cloud-edge cooperation method, identifies the equipment state by using the intelligent analysis algorithm based on collected data, can also be called equipment fault diagnosis, and gives an alarm or/and operates and maintains equipment according to an identification result or a diagnosis result.

Specifically, in this embodiment, the remote robot receives a task sequence issued by the routing inspection scheduling subsystem of the online intelligent inspection system of the substation, starts a task routing inspection function, and sequentially acquires corresponding temperature data, audio data, and appearance data (including meter reading, classification and state, damage, oil leakage, appearance abnormality, and the like) of the equipment in the substation.

And after data are acquired, real-time data processing and equipment fault diagnosis and analysis are respectively carried out, if suspected fault equipment exists, the data enter an intelligent diagnosis expert cloud system for further analysis and processing, then the data enter an item system, the current data acquisition and analysis are subjected to item unified summary processing, and the items are pushed to a UI client side in a corresponding mode.

And (3) task ending process: the condition of task ending is that the current task sequence is executed completely, or abnormally terminated, or the user stops. Entering an item system and reporting items of task ending; and entering an inspection report, performing relevant report processing on the inspection operation, if the operation has suspected fault equipment, entering a suspected fault diagnosis system for further analysis, and generating a report according to an analysis result. And the suspected faulty equipment is stored and recorded, and the record is pushed to the corresponding suspected faulty equipment rechecking task of the user. This concludes the task.

In the above process, as shown in fig. 6, the process flow of generating, executing and subsequently analyzing the inspection task includes the following steps:

step 1: and performing task customization.

Step 2: and selecting the equipment point positions to be inspected by the inspection task.

And step 3: and setting inspection task execution types including comprehensive inspection, routine inspection, light-off inspection, special inspection and special inspection.

And 4, step 4: and setting a task execution period, and executing the inspection task at regular time according to the task period by the inspection task.

And 5: and judging whether the task period is expired, and if so, executing the step 6.

Step 6: and generating a video inspection subtask and a robot inspection subtask through task scheduling.

And 7: the multi-optical device and the multiple robots start concurrent inspection.

And 8: and sending the acquired inspection data file to a pattern recognition service for data recognition to obtain an inspection result.

And step 9: and sending the inspection result to a user interface for real-time display.

Step 10: and carrying out alarm analysis on the inspection result according to the alarm threshold setting of the equipment.

Step 11: and displaying the alarm result through the user interface.

Step 12: and the video polling subtask is ended, and the robot polling subtask is ended.

Step 13: and generating a polling task report.

The process of determining the intelligent analysis algorithm of the station side by using the cloud-edge cooperation method, as shown in fig. 7, specifically includes:

step a: and loading an intelligent analysis algorithm to be updated and an online intelligent patrol system list to be updated.

Step b: and constructing and sending an identity authentication request.

Step c: and analyzing the identity authentication request to perform identity authentication.

Step d; and if the authentication is passed, constructing an intelligent analysis algorithm updating command (the command comprises an authentication legal identification).

Step e: the algorithm data is divided into packets, meanwhile, the sub-packet data is encrypted (the encryption algorithm supports algorithms such as RSA, DES, 3DES, IDEA, MD5 and the like, the algorithm type selection supports configuration), each data packet is added with the total number of the sub-packets, the packet number, the command type and the authentication legal identification to form a command packet, and the command packet is respectively sent to the authentication server of the corresponding online intelligent patrol system according to the online intelligent patrol system list in the update command. Before sending, whether the online intelligent patrol system is online needs to be judged, if the online intelligent patrol system directly sends a command packet, otherwise, a disconnection retransmission mechanism is started.

And the authentication server of the online intelligent patrol system analyzes the update command, judges whether the command type and the authentication legal identification are valid or not, and forwards the command type and the authentication legal identification to the intelligent algorithm analysis unit if the command type and the authentication legal identification are valid. Otherwise, the process ends.

Step f: and after receiving the updating command, the intelligent algorithm analysis unit analyzes the command content and decrypts the algorithm data packet. And after all algorithm data packets are received, verifying whether the algorithm is available, and if the algorithm is available, taking out the old algorithm in the intelligent analysis algorithm bin for compression backup according to the type of the algorithm application object. And adding the algorithm to be updated into the intelligent analysis algorithm bin.

The above-mentioned linkage specifically includes station-side main and auxiliary linkage and station-side environmental system linkage. The following are detailed individually.

The specific scheme of the main-auxiliary linkage at the station end, as shown in fig. 8 or 9, includes:

step I: and the main and auxiliary equipment monitoring system sends the main and auxiliary equipment linkage signals to the patrol processing system through the network safety protection equipment.

Step II: and the online intelligent patrol service in the patrol processing system finds the equipment point location list according to the main equipment code of the linkage signal.

Step III: the online intelligent patrol service acquires a related linkage strategy list from linkage strategy configuration according to the equipment point position list;

step IV: the online intelligent patrol service judges whether the linkage policy list is empty. If the linkage strategy list is not empty, executing the step V; if the linkage strategy list is empty, step XIII is executed.

Step V: triggering linkage.

Step VI: and the online intelligent patrol service constructs a substation equipment patrol task according to the equipment point location list and the linkage strategy list.

Step VII: and starting the patrol task of the substation equipment.

Step VIII: the multi-optical device and the multiple robots start concurrent inspection. The multi-optical equipment concurrent patrol inspection operation is controlled and completed by a video execution unit of the patrol inspection processing system; the multi-robot concurrent inspection is independently completed by a plurality of robots, and the inspection flow of each robot is controlled and completed by a robot execution unit of the robot body system.

Step VIIII: and executing linkage action according to the linkage strategy associated with the equipment point location.

The linkage strategy refers to the combination configuration of various linkage actions and the parameter configuration of the linkage actions.

The linkage action includes: (1) the user can watch the equipment through the real-time video of the user interface of the patrol processing system, and the watching is finished when the watching time exceeds the watching duration configured by the linkage strategy; (2) acquiring, namely acquiring data by the optical equipment or the robot according to an acquired data type configured by the linkage strategy, wherein the acquired data type comprises the following steps: visible light pictures, infrared pictures, visible light videos, infrared videos, audios and partial discharge detection data; (3) collecting and identifying, namely carrying out data collection and then carrying out data identification on the collected data; (4) viewing and collecting, namely, the equipment point location simultaneously executes two linkage actions of viewing and collecting; (5) and (4) watching, acquiring and identifying, namely, simultaneously executing three linkage actions of watching, acquiring and identifying at the point position of the equipment.

Step X: and (5) equipment point location inspection data and inspection result display. The inspection data and the inspection result of the optical equipment are sent to the online intelligent inspection service for storage by the video execution unit of the inspection processing system, the inspection data and the inspection result of the robot are sent to the online intelligent inspection service for storage by the robot execution unit of the robot body system, and then the inspection result and the inspection data are sent to the user interface for display by the online intelligent inspection service.

Step XI: judging whether equipment point positions to be detected exist, namely judging whether the optical equipment passes through a video execution unit of a patrol processing system by the optical equipment, judging whether the robot passes through a robot execution unit of a robot body system by the robot, and executing the step XII when the optical equipment and the robot do not have the equipment point positions to be detected; and if the equipment points to be detected still exist, executing the step VIII.

Step XII: and the on-line intelligent patrol service generates a task report and sends the task report to a user interface of the patrol processing system for display.

Step XIII: and finishing the linkage.

The linkage process of the environmental system at the station end, as shown in fig. 9, includes the following steps:

step a: the method comprises the steps that a user starts a substation equipment inspection task through the user interface control of an inspection processing system or the online intelligent inspection service timing control, wherein the inspection task comprises an inspection task code, an inspection task name, a substation equipment point location list needing inspection, an inspection task priority, an inspection task timing period, an inspection task creator and inspection task creation time.

Step b: the online intelligent inspection service judges whether the equipment point location list of the inspection task contains the protected indoor equipment, if so, the step c is executed; if not, step f is performed.

Step c: the on-line intelligent patrol service sends a control command to the environment monitoring system, and the environment monitoring system controls the lighting controller to turn on the lighting lamp and controls the curtain controller to turn off the curtain. Wherein, turn on the light and prevent to protect indoor light not enough and influence visible light detection, close the (window) curtain and prevent that sunlight shines into and protects indoor influence visible light detection and infrared detection.

Step d: the online intelligent inspection service judges whether the temperature of the protection room exceeds a threshold value through the temperature of the protection room acquired by the environment monitoring system, and if so, the step e is executed; if not, step f is performed.

Step e: the online intelligent patrol service sends a control command to the environment monitoring system, the environment monitoring system controls the air conditioner controller to open the air conditioner, and the temperature of the protection room is adjusted to a normal range to prevent the influence on infrared detection.

Step f: and the online intelligent inspection service carries out task scheduling to generate a video inspection subtask and a robot inspection subtask.

Step g-A: and a video execution unit of the patrol processing system executes a video patrol subtask.

Step h-A: and the video execution unit controls the optical equipment to patrol the equipment point location.

Step i-A: if the current equipment point is outdoor equipment, the current optical equipment is deployed outdoors, a video execution unit of the patrol processing system judges whether rain and snow exist currently through a rain and snow sensor of the environment monitoring system, and if so, the step j-A is executed; if not, step k-A is performed.

Step j-A: the video execution unit controls the optical apparatus to perform a wiper operation.

Step k-A: the video execution unit controls the optical equipment to collect the polling data of the equipment point location, the polling data is sent to the mode identification service of the polling processing system to complete data identification, a polling result is obtained, then the polling data and the polling result are sent to the online intelligent polling service, the online intelligent polling service completes alarm analysis, and the online intelligent polling service sends the polling result and the alarm to the user interface of the polling processing system to perform polling result and alarm display.

Step l-A: the video execution unit judges whether the equipment point location to be detected still exists, if so, the step h-A is executed; if not, step u is performed.

Step g-B: and the robot execution unit of the robot body system executes the robot inspection subtask.

Step h-B: and the robot execution unit controls the robot to patrol the point location of the equipment.

Step i-B: the robot execution unit judges whether the robot switches indoor and outdoor environments, and if so, executes the step j-B; if not, step k-B is performed.

Step j-B: the robot execution unit sends a control command to the environment monitoring system, the environment monitoring system controls the automatic door controller to open the automatic door of the protection room, and the automatic door of the protection room is closed after the robot enters or leaves the protection room.

Step k-B: if the current equipment point is outdoor equipment, the current robot runs outdoors, the robot execution unit judges whether rain and snow exist currently through a rain and snow sensor of the environment monitoring system, and if so, the step l-B is executed; if not, step p-B is performed.

Step l-B: the robot execution unit acquires rainfall and snowfall through a rain and snow sensor of the environment monitoring system, judges whether the current rainfall and snowfall exceed a threshold borne by the robot or not, and executes the step n-B and the step q-B if the current rainfall and snowfall exceed the threshold borne by the robot; if not, step m-B is performed.

Step m-B: the robot executing unit controls the robot to perform a wiper operation and then performs the step p-B.

Step n-B: and the online intelligent patrol service judges whether the current equipment point location is associated with optical equipment, and if so, the step o-B is executed.

Step o-B: and the online intelligent inspection service sends a control command to the video execution unit, the video execution unit controls the optical equipment to execute wiper operation, controls the optical equipment to inspect the current equipment point position, and then executes the step p-B.

Step p-B: the robot execution unit controls the robot to collect the polling data of the equipment point location, the polling data is sent to the mode identification service of the robot body system to complete data identification, a polling result is obtained, the polling data and the polling result are sent to the online intelligent polling service, the online intelligent polling service completes alarm analysis, and the online intelligent polling service sends the polling result and the alarm to the user interface of the polling processing system to perform polling result and alarm display. Step t-B is then performed.

Step q-B: the robot execution unit sends the current position and the position of the charging point of the robot to a path planning service of the robot body system, the path planning service sends a calculated route of the robot returning to the charging room to the robot execution unit, and the robot execution unit controls the robot returning to the charging room according to the route.

Step r-B: the robot execution unit controls the robot to enter a charging state and wait.

Step s-B: the robot execution unit acquires rainfall and snowfall through a rain and snow sensor of the environment monitoring system, judges whether the current rainfall and snowfall are lower than a threshold value born by the robot or not, and executes the step t-B if the current rainfall and snowfall are lower than the threshold value born by the robot; if not, step r-B is performed.

Step t-B: the robot execution unit judges whether a point position of the equipment to be detected still exists, and if yes, the step h-B is executed; if not, step u is performed.

Step u: and finishing the polling task.

Step v: and the online intelligent inspection service generates an inspection task report and sends the inspection task report to a user interface of the inspection processing system for display.

The inspection task includes each inspection point that the robot needs to inspect, and in this embodiment, the configuration flow of the inspection points, as shown in fig. 10, includes:

acquiring a three-dimensional model of a polling place, and preprocessing the three-dimensional model;

and calculating the corresponding three-dimensional posture of the inspection mechanism according to each target object and the corresponding optimal three-dimensional coordinates of the inspection point to obtain an inspection point information list of the inspection mechanism.

Specifically, the specific process of preprocessing the three-dimensional model includes: and carrying out noise point removal and feature fitting on the three-dimensional model.

The specific process of identifying the target object in the inspection place comprises the following steps: and loading a preset training model base, identifying the target objects in the whole station by adopting a feature matching method, and calculating the pose, the orientation normal vector and the size information of each identified target object in the three-dimensional model.

The distance range between the inspection point and the target object is determined according to the target object to be detected, the maximum distance and the minimum safety distance between the inspection mechanism and the target object are determined theoretically when the inspection mechanism executes inspection, and the distance range of the preset inspection point is determined based on the maximum distance and/or the minimum safety distance.

In this embodiment, a sphere equation can be established according to the theoretical maximum distance between the preset inspection point and the target object, and a cross line of a plane where the sphere and the passable path are located is solved, where the cross line is an outer ring boundary of the legal configuration area; according to the constraint of the maximum included angle between the central line of the camera field of view and the normal line of the target object plane, establishing a cone equation, solving a cross line of a cone and the plane where the passable path is located, wherein the cross line is the inner circle boundary of a legal configuration area, two cross lines are crossed in an arc line, in the cross area, calculating the boundary of the passable path falling into the cross area, and all passable path areas meeting the boundary interval are legal configuration areas corresponding to a single target object.

In the public crossing area, a weighted cost optimization function of the distances and deviation angles of a plurality of target objects is established, the cost function is weighted according to the distance cost of the plurality of target objects and a preset patrol point, the included angle cost of a preset patrol point camera view field central line and a plurality of target object plane methods, the distance cost and the included angle cost, and an optimization method is adopted to solve the optimal solution of the cost function, namely the optimal solution is the preset patrol point of the multiplexing crossing area of the plurality of target objects.

For target objects which do not have intersection in a legal configuration area, establishing a weighted cost optimization function of the distance and the deviation angle of a single target object, wherein the cost function comprises the distance cost of the single target object and a preset inspection point, and the cost of an included angle between the central line of the field of view of a camera at the preset inspection point and the normal line of the plane of the single target object: and weighting the distance cost and the included angle cost, and solving the optimal solution of the cost function by adopting an optimization method, namely the optimal solution is the special preset inspection point for the single target object.

During conversion, according to each target object and the corresponding preset three-dimensional coordinates of the patrol point, the three-dimensional posture of the holder is calculated through a trigonometric function relationship, firstly, the coordinates of the rotation central point of the holder are solved at the preset coordinates of the patrol point according to the robot body model, then, the trigonometric function relationship is established according to the coordinates of the target object and the coordinates of the rotation central point of the holder, and the three-dimensional posture of the holder is solved.

Through the process of routing inspection, the running state of the equipment is comprehensively identified based on the video information, the appearance data of the target equipment and the sound information of the target equipment, and when at least one information analysis result is that a fault occurs, the fault is determined to occur.

As shown in fig. 12, the process of performing state identification based on voiceprint information includes:

acquiring sound information of target equipment of a transformer substation, and preprocessing the sound information;

The process of extracting the FBank features comprises the following steps:

obtaining a spectrogram of the sound signal through Fourier transform, and calculating an energy spectrum;

filtering through a Mel filter bank to obtain a sound spectrum which accords with the hearing habits of human ears;

and solving the natural logarithm of the sound spectrum obtained after filtering to obtain the FBank characteristic.

According to the inspection result, the equipment operation and maintenance can be performed, and the specific process is shown in fig. 11, and the method comprises the following steps:

step 1-A: and drawing up a routine operation and maintenance plan through a user interface by a user, wherein the routine operation and maintenance plan comprises a transformer station name, a routine operation and maintenance plan name, a plan starting time, a plan ending time and a plan execution period.

Step 1-B: the user checks routine operation and maintenance items on time, and finds and records equipment inspection problems;

step 2-A: routine patrol tasks are performed normally.

Step 2-B: in the patrol process, patrol equipment (intelligent robots or optical equipment) generates an abnormal alarm.

And step 3: and judging whether the inspection equipment problems found by routine operation and maintenance project inspection or inspection equipment abnormity alarms generated in the normal inspection process can be automatically solved by the user.

Step 3-A: if the user can solve the problem by himself, the next routine operation and maintenance of the user is normally carried out or routine inspection tasks are normally executed.

Step 3-B: if the user can not solve the problem by himself, the problem is fed back to an equipment manufacturer for professional maintenance; the personnel of the manufacturer actually solve the problem of inspecting the equipment on site and remove the fault; and recording the professional operation and maintenance condition for subsequent filing and retrieval.

Example three: a computer readable storage medium, wherein a plurality of instructions are stored, the instructions are suitable for being loaded by a processor of a terminal device and executing the steps of the substation online intelligent patrol method mentioned in the second embodiment.

Example four: a terminal device comprising a processor and a computer readable storage medium, the processor being configured to implement instructions; the computer readable storage medium is used for storing a plurality of instructions, and the instructions are suitable for being loaded by the processor and executing the steps of the substation online intelligent patrol method mentioned in the second embodiment.

As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Although the embodiments of the present invention have been described with reference to the accompanying drawings, it is not intended to limit the scope of the present invention, and it should be understood by those skilled in the art that various modifications and variations can be made without inventive efforts by those skilled in the art based on the technical solution of the present invention.